Heparin and Venoarterial Extracorporeal Membrane Oxygenation in Massive Pulmonary Embolism Management: A Case Report

Introduction

Pulmonary embolism is often caused by thrombi primarily composed of fibrin and red blood cells dislodging from the systemic circulation, leading to pulmonary artery occlusion. Approximately 70% to 80% of pulmonary embolism cases originate from deep vein thrombosis in the lower extremities or pelvis, while approximately 6% arise from deep veins in the upper extremities [1]. The clinical presentation demonstrates considerable heterogeneity, influenced by the extent of embolism and the presence of underlying chronic cardiopulmonary disease. Globally, the annual incidence of pulmonary embolism ranges from 39 to 115 cases per 100 000 individuals, with a mortality rate of 9.4 to 32.2 per 100 000 [2]. Diagnosis relies on an integrated approach that combines pre-test probability, D-dimer testing, and imaging to facilitate comprehensive assessment and stratified management. Evaluating the hemodynamic status is essential for differentiating between high-risk and non-high-risk pulmonary embolism. The Pulmonary Embolism Severity Index and its simplified version (sPESI) are validated prognostic tools that effectively stratify patients based on their 30-day mortality risk [3]. Treatment strategies vary depending on risk classification and may include anticoagulation, systemic thrombolysis, catheter-directed thrombolysis (CDT), catheter thrombectomy, catheter-assisted fragmentation techniques, and surgical thrombectomy. The case presented in this report illustrates a complex scenario, with simultaneous high thrombotic and bleeding risks for which no clear evidence-based guidance or academic references were available. Through repeated resuscitative efforts and careful balancing of therapeutic options, successful treatment was achieved. Treatment of this case may provide valuable insights for future guideline development and standardized care.

Case Report

A 61-year-old woman with a history of hypertension (baseline blood pressure approximately 130/80 mmHg; 1 mmHg=0.133 kPa) and several years of unexplained right calf discomfort and pain, for which she had not sought treatment, was admitted to a local hospital on April 14, 2025, at 09: 00. She presented with persistent right axillary and anterior chest wall pain accompanied by dyspnea for 6 days. An electrocardiogram (ECG; Figure 1) showed ST-segment depression in leads I, aVL, and V6, accompanied by atrioventricular block, suggestive of acute myocardial infarction (AMI). She was administered 300 mg of aspirin and 600 mg of clopidogrel bisulfate orally and was scheduled for emergency percutaneous coronary intervention. However, preoperative laboratory results returned with normal myocardial infarction biomarkers, which did not support AMI, prompting her transfer to our hospital. By 11: 00, upon arrival at our emergency department, a repeat ECG revealed an SIQIIITIII pattern (Figure 2). Doppler echocardiography demonstrated a D-sign (Figure 3), a tricuspid annular plane systolic excursion (TAPSE) of 0.99 mm, a right ventricular/left ventricular (RV/LV) ratio of 0.73, and a pulmonary artery pressure of 39.36 mmHg. Ultrasonography confirmed thrombosis in the right popliteal vein, muscular veins, and superficial branch of the peroneal vein. During the evaluation, the patient experienced sudden syncope lasting approximately 5 seconds. Consequently, computed tomography pulmonary angiography (CTPA) was not performed, and she was admitted to the intensive care unit (ICU) with a provisional diagnosis of pulmonary embolism.

Upon ICU admission, the patient was conscious and coherent, with alleviated chest pain and dyspnea compared to the morning. Hemodynamically stable without vasopressor support, her blood pressure fluctuated around 90/60 mmHg, with a heart rate of 104 beats per minute. Arterial blood gas analysis under high-flow nasal cannula oxygen therapy (FiO2 0.55, flow rate 45 L/min) revealed a partial pressure of oxygen of 74 mmHg. Repeat laboratory tests indicated hypercoagulability, elevated troponin and N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels, and mild abnormalities in hepatic and renal function (Table 1). Based on the presentation of syncope, hypotension, and deep vein thrombosis, the patient could be stratified as having high-risk pulmonary embolism. However, the potential for high bleeding risk with systemic thrombolysis in the context of dual antiplatelet therapy (DAPT) was a significant concern. This was corroborated by a “Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes with Early Implementation of the ACC/AHA Guidelines?” (CRUSADE) bleeding score of 58, placing the patient in a high-risk category for subsequent antithrombotic therapy. Given this baseline profile of simultaneously high thrombotic burden and high bleeding risk, conservative anticoagulation was selected as the initial therapeutic strategy after discussion with the family. The specific regimen consisted of an intravenous bolus of unfractionated heparin at 80 IU/kg, followed by a continuous infusion at 18 IU/kg/h, with a target activated partial thromboplastin time (APTT) ratio of 1.5 to 2.5 times the baseline control.

At 22: 00, the patient experienced worsening chest pain and dyspnea, and the latest coagulation monitoring showed an APTT of 72.9 seconds. Non-invasive ventilation and analgesic symptomatic treatment were administered but provided no significant relief. Subsequently, the patient developed hypotension and was started on a continuous infusion of norepinephrine bitartrate at 0.18 μg/kg/min (patient weight: 75 kg). Repeat arterial blood gas analysis and laboratory tests revealed metabolic acidosis, hepatic and renal dysfunction, and a significant increase in myocardial injury markers compared with previous levels (Table 1). Color Doppler echocardiography demonstrated a thrombus in the main pulmonary artery (Figure 4), a TAPSE of 0.89 mm, an RV/LV ratio of 1.2, and a pulmonary artery pressure of 56 mmHg.

By approximately 23: 00, the patient’s consciousness had deteriorated to lethargy, and the vasopressor requirement increased to 1.42 μg/kg/min. The patient was sedated, analgesized, endotracheally intubated, and placed on mechanical ventilation. Emergent thrombolytic therapy was initiated following the cessation of heparin. Due to the patient’s prolonged APTT and a clinical presentation complicated by disseminated intravascular coagulation (DIC) and evolving multiorgan dysfunction, the standard dose of 100 mg alteplase presented a significant risk of bleeding. Given evidence that a half-dose regimen of alteplase maintains considerable efficacy, a cautious approach was adopted, administering a reduced dose of 50 mg alteplase (Actilyse) infused over 2 hours. Unfortunately, no clinically meaningful improvement in shock was observed after 4 hours, and laboratory parameters continued to deteriorate.

A multidisciplinary expert discussion was convened, concluding that interventional thrombectomy with inferior vena cava filter placement under VA-ECMO support might be a superior treatment strategy. For venous drainage, a 17-F cannula was selected. Considering the potential need for subsequent inferior vena cava filter placement, repeat catheter-based interventions, or even intra-aortic balloon pump support in the event of severe cardiogenic shock, the left femoral arterial and venous access was deliberately preserved. Therefore, venous cannulation was performed via the right internal jugular vein, with the cannula tip advanced into the right atrium to ensure adequate drainage flow. A 15F return cannula was placed in the right femoral artery. By 09: 00 the next day, VA-ECMO was successfully established and running at a flow rate of 3.5–4.0 L/min, leading to gradual hemodynamic improvement. Follow-up echocardiography showed a TAPSE of 1.35 mm, an RV/LV ratio of 0.9, and a pulmonary artery pressure of 38 mmHg.

The patient was then transferred to the interventional suite. Using a modified Seldinger’s technique, the right femoral vein was punctured, and a 5F pigtail catheter was inserted for pulmonary angiography. The angiogram revealed a continuous, strip-like filling defect straddling the left and right main pulmonary arteries, with occlusion of the left upper pulmonary artery (Figure 5). An 8F sheath and an 8F thrombus aspiration catheter were subsequently introduced; however, only a small amount of thrombotic material was aspirated, and the filling defect in the main arteries showed no significant change. This raised suspicion regarding the nature of the obstruction, with differential diagnoses including pulmonary artery sarcoma or a foreign body. An inferior vena cava filter was placed. The patient then underwent CTPA in the CT suite (Figure 6), which still indicated densities consistent with thrombus.

The remote specialist consultation recommended continuing unfractionated heparin anticoagulation with close monitoring and prompt transfer to a higher-level hospital if clinical deterioration occurred. Anticoagulation was therefore resumed when the APTT fell below twice the baseline value, maintaining it at approximately 1.5 times the baseline. Fortunately, the patient’s subsequent clinical course was favorable, complicated only by minor vaginal bleeding. Repeat CTPA performed 3 days later showed thrombus regression (Figure 7). During this period, all monitored parameters gradually improved, and ECMO support was progressively weaned. Successful decannulation from ECMO was achieved on day 7. Three weeks later, the patient regained good neurological and motor function. A follow-up CTPA (Figure 8) demonstrated significant resolution of the residual thrombus. However, due to persistent oliguria and a creatinine clearance of 18.6 mL/min, she required ongoing intermittent continuous renal replacement therapy and was subsequently transferred to a specialized nephrology hospital.

Discussion

This case describes a patient with chronic venous thromboembolism who lacked typical predisposing factors and subsequently developed an acute pulmonary embolism. Two weeks after the primary event, laboratory assessment for thrombophilia while on low-molecular-weight heparin treatment indicated decreased levels of antithrombin III (49%), protein C activity (44%), and protein S activity (53.7%). The presence of heparin could have influenced the results, preventing a conclusive diagnosis of hereditary thrombophilia. Nevertheless, the results suggested the presence of severe hereditary conditions like antithrombin deficiency or combined thrombophilia. Regrettably, confirmatory genetic disorder whole-exome sequencing was not conducted. High-risk pulmonary embolism is currently defined as systolic blood pressure <90 mmHg for more than 15 minutes, a ≥40 mmHg drop from baseline, or the need for vasopressor support [4]. In patients without high bleeding risk, systemic thrombolysis is considered an emergency treatment indication. For patients with high bleeding risk or those in whom thrombolysis fails, current guidelines recommend CDT and catheter thrombectomy as rescue treatments [3–5], with CDT often preferred [4]; in certain cases, the 2 treatments can be combined. However, there remains a lack of randomized controlled trials comparing systemic thrombolysis and local thrombolysis in high- and intermediate-risk patients with pulmonary embolism, and reported rates of major bleeding vary considerably among studies [3], with some reports suggesting a rate comparable to that of systemic thrombolysis [6]. Large-bore aspiration thrombectomy has emerged as a potentially beneficial option for patients at a high risk of bleeding, owing to its rapid thrombus clearance and the avoidance of thrombolytic agents [7]. However, its technical demands, the risk of procedure-related complications, and the absence of clear guideline endorsements as a first-line therapy in all situations have prevented advanced systems such as FlowTriever and Indigo from achieving widespread global adoption [8]. The present case, which illustrates the suboptimal efficacy of conventional catheters against massive thrombi, highlights the need to equip regional hospitals with specialized pulmonary embolism thrombectomy systems to effectively address the increasing burden of this disease.

In the current case, initial anticoagulation therapy was administered mainly for the following reasons: (1) The patient’s symptoms improved after loading with DAPT. Experimental studies have confirmed that platelets aggregate at the “head” of venous thrombi early in formation. Aspirin has been shown to improve tachycardia, pulmonary hypertension, and systemic hypotension in patients with pulmonary embolism by inhibiting activated platelet release of prostaglandins and serotonin [9]. (2) Based on bleeding risk assessment during antiplatelet therapy for acute coronary syndrome, the CRUSADE score was 58, indicating very high bleeding risk, and the patient was at peak plasma concentration of DAPT upon ICU admission. (3) Although neither absolute nor relative contraindications to systemic thrombolysis in pulmonary embolism explicitly mention DAPT loading, there are no guidelines or publications that clearly state that it is safe to proceed to thrombolysis following DAPT loading. Notably, there is consensus among professional societies that immediate empiric therapeutic anticoagulation should be initiated in patients with intermediate or high clinical suspicion of pulmonary embolism and low bleeding risk, particularly in those suspected of having high-risk pulmonary embolism, even while confirmatory testing is pending [10].

The European Society of Cardiology/European Respiratory Society guidelines recommend prompt administration of body weight-adjusted loading doses of heparin. However, when patients worsen after anticoagulation, the guidelines support thrombolysis in only non-high-risk patients and prefer CDT as the first-line approach [11]. In our case, the patient rapidly developed refractory shock and metabolic acidosis due to hemodynamic deterioration. Considering that CDT requires invasive catheter placement and patient transfer, systemic intravenous thrombolysis was considered a faster and potentially safer approach at the time of treatment. Unfortunately, despite continuous infusion of 50 mg of recombinant tissue plasminogen activator (rt-PA) for over 2 h, the patient experienced no significant improvement, only stabilization of hemodynamics without further deterioration. The analysis of potential causes for failure may encompass the substantial thrombus burden and the delay in intervention that extended beyond the optimal thrombolytic window of 48 hours, given that the onset occurred 6 days prior. Additionally, considering the patient’s history of chronic deep venous thrombosis and the intraoperative observation of thrombus morphology, the presence of chronic organized components cannot be dismissed, indicating a mixed acute-on-chronic embolism. Therefore, it can be inferred that CDT, which typically utilizes only one-third to one-fourth of the systemic thrombolytic dose, would likely have demonstrated even lower efficacy under these conditions. Furthermore, clinical trials [12,13] have demonstrated that a reduced dose of rt-PA (50 mg/2 h) maintains comparable efficacy to the conventional standard dose of rt-PA (100 mg/2 h) while significantly decreasing the risk of bleeding. This regimen is especially appropriate for Asian populations, in which the average body weight is typically lower, and it offers robust evidence-based support for optimizing thrombolytic strategies for pulmonary embolism on a global scale.

Unexpectedly, our patient demonstrated improvement with anticoagulation therapy while supported by VA-ECMO, although the total duration of circuit use exceeded the median of 5.1 days reported in previous studies [7]. This extended duration may be linked to persistently low levels of antithrombin III during treatment, which could have compromised the antithrombotic efficacy of heparin. Thrombocytopenia is relatively prevalent among ECMO patients, with prior studies identifying causes such as operative or spontaneous bleeding, platelet activation-related injury due to contact with artificial materials and mechanical shear stress, splenic sequestration, and diagnostic phlebotomy [14]. Notably, platelet activation is responsible for approximately 32% of overall platelet loss. Furthermore, it has been observed that patients with pre-existing DIC or severe bleeding during extracorporeal life support experience a doubled rate of platelet loss [15]. In this case, the significant decrease in platelet count prior to cannulation aligned with the characteristics of DIC.

Finally, the improvement of the patient with anticoagulation therapy under VA-ECMO support was unexpected. Current guidelines recommend VA-ECMO in high-risk pulmonary embolism primarily for refractory shock or cardiac arrest or as a bridge to catheter-based or surgical interventions [16]. However, data suggest its use in pulmonary embolism treatment is becoming increasingly common [16]; numerous studies have evaluated the risks and benefits of VA-ECMO alone or in combination with thrombolysis, catheter-based interventions, or surgery, although the findings remain inconsistent [16–18]. A 2021 retrospective study from Geneva University Hospitals [19] reviewed 36 patients with obstructive and/or refractory cardiogenic shock caused by massive acute pulmonary embolism (MAPE) over 10 years. Among the 17 patients treated with fibrinolysis and/or catheter thromboaspiration while on VA-ECMO (16 received fibrinolysis and 5 catheter-directed thromboaspiration), an overall survival rate of 64% was achieved, significantly better than the baseline mortality of up to 50% for MAPE. However, initiating VA-ECMO after failing thrombolysis markedly increased bleeding complications and worsened 30-day mortality, with mortality in the VA-ECMO plus thrombolysis group significantly higher than that in the VA-ECMO alone group. In contrast, an emulated target trial by Stadlbauer et al [18] found that managing high-risk pulmonary embolism solely with VA-ECMO without subsequent recanalization was associated with the highest estimated in-hospital mortality. Surgical embolectomy provided the greatest survival benefit, while in-hospital systemic thrombolysis and percutaneous catheter-directed therapy showed modest advantages compared with VA-ECMO alone.

Despite these mixed findings, the overall evidence points toward potential benefits.

Multicenter observational studies have shown that VA-ECMO may reduce in-hospital mortality in selected patients, regardless of whether it is combined with advanced reperfusion therapy [16,20]. Although early initiation of ECMO therapy has been associated with reduced risk of cardiac arrest and in-hospital mortality [21], George et al [22] found no benefit of ECMO in patients with cardiac arrest and blood lactate levels >6 mmol/L [23]. Overall, patients in studies demonstrating favorable outcomes typically initiate ECMO support prior to the onset of irreversible terminal organ damage, whereas those with unfavorable outcomes may include a greater proportion of patients who are already in a dying state or have experienced cardiac arrest. Furthermore, the timing and application of reperfusion methods in conjunction with ECMO represent critical variables influencing final prognosis. This underscores that the efficacy of this strategy relies heavily on careful patient selection and precise timing of initiation, rather than on ECMO as a treatment modality.

Furthermore, it should be emphasized that the efficacy in this case is primarily attributable to the hemodynamic unloading and organ perfusion support provided by VA-ECMO, which created a critical time window for the recovery of right ventricular function. The role of heparin lies in preventing the formation of new thrombi within the circulatory system and pulmonary arteries, while it exerts no direct lytic effect on pre-existing, particularly organized, thrombi. This observation suggests that clinicians may consider such a strategy as an important conditional rescue option in scenarios where reperfusion therapy has failed or is contraindicated, or in the presence of high bleeding risk, excessive thrombus burden, or delayed presentation. Under these circumstances, this approach can both prevent thrombus propagation via anticoagulation and buy time for potential spontaneous recanalization or subsequent elective intervention. Compared with catheter-based interventions directly targeting the thrombus, its advantage lies in the immediate reversal of life-threatening circulatory failure, whereas its limitations include not addressing the underlying thrombotic occlusion and carrying the inherent risks associated with ECMO.

Conclusions

This case highlights the complexity of clinical decision-making in cases of high-risk pulmonary embolism. The rapid advancement of catheter-based interventions and mechanical circulatory support devices has further complicated therapeutic choices. Based on the experience from our case and the existing body of evidence, we advocate for the development of a unified assessment framework to guide the potential selective strategy of earlier anticoagulation therapy with ECMO support. This framework could incorporate a combination of parameters such as the requirement for moderate-dose vasopressors (eg, norepinephrine >0.3μg/kg/min), elevated lactate levels (>4 mmol/L), objective measures of right ventricular dysfunction (eg, RV/LV diameter ratio >1.2, TAPSE <12 mm), assessment of thrombus burden and characteristics, failed aspiration attempts, and comprehensive bleeding risk evaluation. The goal is to systematically identify high-risk pulmonary embolism patients for whom conventional reperfusion therapy is failing, unavailable, or contraindicated and who are progressing toward circulatory collapse. Furthermore, the establishment of standardized guidance for intra-procedural anticoagulation targets, therapeutic response evaluation, and ECMO weaning criteria is essential to better inform clinical implementation. Hence, this case underscores the increasing importance of ECMO in managing pulmonary embolism, revealing substantial gaps in understanding regarding the optimal timing for initiation, patient selection criteria, and the precise timing and approach for integrating it with delayed catheter-directed interventions. The effectiveness implied by this case is merely a preliminary hypothesis requiring validation. The overall efficacy and safety of this approach must be established through prospective, multicenter controlled trials. Its primary value lies in reframing a complex clinical issue into a set of clear, testable scientific inquiries. It is designed to spark and direct future research endeavors aimed at addressing these knowledge voids, ultimately facilitating the delivery of more evidence-based, personalized treatment strategies to high-risk patients in this ambiguous therapeutic realm.